Real-Time Simulation Of Dust Behavior Generated By A Fast .
Real-Time Simulation of Dust BehaviorGenerated by a Fast Traveling VehicleJIM X. CHEN, XIAODONG FU, and EDWARD J. WEGMANComputer Graphics Laboratory, George Mason UniversitySimulation of physically realistic complex dust behavior is very useful in training, education,art, advertising, and entertainment. There are no published models for real-time simulation ofdust behavior generated by a traveling vehicle. In this paper, we use particle systems,computational fluid dynamics, and behavioral simulation techniques to simulate dust behavior in real time. First, we analyze the forces and factors that affect dust generation and thebehavior after dust particles are generated. Then, we construct physically-based empiricalmodels to generate dust particles and control the behavior accordingly. We further simplifythe numerical calculations by dividing dust behavior into three stages, and establishingsimplified particle system models for each stage. We employ motion blur, particle blending,texture mapping, and other computer graphics techniques to achieve the final results. Ourcontributions include constructing physically-based empirical models to generate dust behavior and achieving simulation of the behavior in real time.Categories and Subject Descriptors: I.6.5 [Simulation and Modeling]: Model Development—Modeling methodologiesGeneral Terms: Simulation, Graphics, DustAdditional Key Words and Phrases: Physically-based Modeling, Real-time Simulation, Vehicle,Particle Systems, Computational Fluid Dynamics1. INTRODUCTIONIn many virtual environments and distributed interactive simulations, it isdesirable to simulate trucks, armored vehicles, bulldozers, and otherground-based moving objects [Balci 1997; Chen et al. 1997; Cremer et al.1995; Schiavone et al. 1997; IST 1994; Li and Moshell 1993]. Typically,The work of Dr. Wegman was supported by the Army Research Office under contractDAAH04-94-G-0267 and by a National Science Foundation Group Infrastructure GrantDMS-9631351.Authors’ addresses: J. X. Chen and X. Fu, Computer Graphics Laboratory, George MasonUniversity, Fairfax, VA 22030; email: jchen6@gmu.edu; xfu@gmu.edu; E. J. Wegman, Centerfor Computational Statistics, George Mason University, Fairfax, VA 22030; email:ewegman@galaxy.gmu.edu.Permission to make digital / hard copy of part or all of this work for personal or classroom useis granted without fee provided that the copies are not made or distributed for profit orcommercial advantage, the copyright notice, the title of the publication, and its date appear,and notice is given that copying is by permission of the ACM, Inc. To copy otherwise, torepublish, to post on servers, or to redistribute to lists, requires prior specific permissionand / or a fee. 2000 ACM 1049-3301/99/0400 –0081 5.00ACM Transactions on Modeling and Computer Simulation, Vol. 9, No. 2, April 1999, Pages 81–104.
82 J. X. Chen et al.dust behavior is not simulated when these objects travel on an unpavedroad. Dust behavior is caused by different factors (such as natural windand a fast traveling vehicle) and appears in many situations. Simulatingphysically realistic, complex dust behavior is very useful in interactivegraphics applications, such as computer art, advertising, education, entertainment, and training. However, due to the lack of modeling and simulation techniques, there are currently no published successful real-timesimulations of realistic dust behavior. As computers and their graphicssystems become faster, many natural phenomena (such as the behaviors offluids, terrains, trees, fireworks, volcanos, clouds, falling leaves, etc.) maybe simulated in real time [Chen and Lobo 1995; Li and Moshell 1993; Lokeet al. 1992; Oppenheimer 1986; Reeves 1983; Reynolds 1987; Roth andGuritz 1995; Sims 1990; Stam and Fiume 1993; Wejchert and Haumann1991]. We believe it is appropriate now to include dust behavior intoreal-time simulation.Hsu and Wong [1995] introduced a dust accumulation model. Their modelpresents static appearance of dust accumulation without behavior andanimation. Cowherd [1989], Williams [Williams and Davis 1988], and otherresearchers studied dust and the mechanisms of dust generation. Theirpurpose was to study and measure the density of the dust in the realbattlefield rather than simulating the dust behavior graphically. Today,military training using graphics and distributed interactive simulation isan important topic of research and applications [IST 1994], and generatingdust behavior in real time can significantly increase the realism of thesimulated training environment.In this paper, we introduce a method for simulating the dust behaviorcaused by a fast traveling vehicle in real time. Specifically, we can calculateand render the dust behavior at more than four frames per second (with2000 particles animated at the same time) on a SGI OnyxII graphicscomputer. The method exploits a combination of particle systems, rigidbody particle dynamics, computational fluid dynamics (CFD), and behavioral simulation techniques. The particle systems technique was first introduced to computer graphics by Reeves [1983], and is now widely used tosimulate fuzzy or dynamic objects, such as fire, grass, explosions, clouds,water, and trees. These objects have no fixed shape and change theirshapes and behavior stochastically. They have ill-defined boundaries thatmake surface-based modeling impractical. Dust behavior behind a movingvehicle belongs to this category. CFD methods are employed to generate thedynamics of air flow behind a moving vehicle. We also employ motion blurfor small and fast-moving particles, particle blending instead of hiddensurface removal, texture mapping, and other graphics techniques toachieve both better performance and also better appearance in the finalresults.In order to construct a physically based realistic simulation, we firstanalyze the forces and factors that affect dust generation and the behaviorafter dust particles are generated. Then we construct physically basedempirical models for generating dust particles and for controlling theACM Transactions on Modeling and Computer Simulation, Vol. 9, No. 2, April 1999.
Real-Time Simulation of Dust Behavior 83behavior. The particle systems are integrated with CFD to achieve betterrealism. However, the models are time-consuming and inefficient. Based onthese general models and the analysis of forces, we simplify the numericalcalculations by dividing dust behavior into three stages, and establishingsimplified particle system models for each stage. The resulting models aresatisfactory for real-time simulation as well as for achieving realistic dustbehavior.The rest of the paper is organized as follows: In Section 2, we brieflydescribe the particle systems technique that is employed in our dustsimulation. In Section 3, we discuss the generation of a dust particle, theturbulent air flow affecting the behavior after a dust particle is generated,and the dynamics of a dust particle. In Section 4, we simplify the models,and divide the dust particle system into three stages (fluid turbulence,particle momentum, and airborne drift). In Section 5, we discuss somerendering issues and present some simulation results with different parameters. Finally, in Section 6, we summarize and describe several avenues offuture research.2. PARTICLE SYSTEMSThe original particle systems method proposed by Reeves [1983] was basedon stochastic processes. Each particle is independent and moves accordingto its own characteristics. He hypothesized no interaction among particlesin a particle system. Structured particle systems, which are used tosimulate objects that are more structured (such as trees, grass, etc.), werealso proposed by Reeves. Sims [1990] extended Reeves’ work and allowedthe independent particles to interact with the environment. Today, thereare numerous publications on particle systems. The interactions amongparticles in a particle system and among particle systems are very important for simulating some phenomena. For example, a particle systemrepresenting fire could interact with another particle system representingwater. This will result in a new particle system representing steam.2.1 PropertiesA particle-based simulation may include a number of independent particlesystems. Each particle system is made of many particles and each particlehas its own parameters that significantly influences the particles’ properties. In general, a particle system and its particles have very similarparameters, but with different values:—Position (including orientation in 3D space and center location x, y, and z)—Movement (including velocity, rotation, acceleration, etc.)—Color (RGB)—Transparency (alpha)—Shape (point, line, sphere, cube, rectangle, etc.)ACM Transactions on Modeling and Computer Simulation, Vol. 9, No. 2, April 1999.
84 J. X. Chen et al.—Volume—Density—Mass—Lifetime (only for particles)—Blur head and rear pointers (only for particles)The position, shape, and size of a particle system determine the initialpositions of the particles and their range of movement. The movements ofthe particles are restricted within the range defined by their associatedparticle system. The shape of a particle system can be a point, linesegment, sphere, box, or cylinder. The movement of a particle system isaffected by internal or external forces, and the results of the rotations andaccelerations of the particles as a whole. A particle system may change itsshape, size, color, transparency, or some other attributes as it evolves.The lifetime defines how many time slices (frames) a particle will beactive. A particle has both a head position and a tail position. The headposition is usually animated and the tail position follows along for motionblur. The simulation looks more dynamic and has more particles withmotion blur, at the cost of longer rendering time.2.2 SimulationIn general, particle systems are first initialized with each particle havingits original position, velocity, color, transparency, shape, size, mass, andlifetime. After the initialization, for each simulation frame, some parameters of the particles are updated using a rule base, and the resultingparticle systems are rendered. An outline of a general simulation loop for aparticle system is as follows:—Create particle systems: ranges of particle activities; data structures tohold particles—Initialize each particle’s parameters (position, velocity, color, transparency, shape)—For each simulation frame, do the following:—Create and remove particles—Calculate and update each particle’s parameters (position, velocity)—Adjust each particle’s motion blur (head and rear) pointers—Render the particles in the particle systems—Update the particle systems’ ranges of activities2.3 ApplicationsFigure 1 summarizes applications using particle systems or particle-basedsimulation in computer graphics.As shown above, a particle system has different shape, color, transparency properties, and movement governed by certain underlying processes.ACM Transactions on Modeling and Computer Simulation, Vol. 9, No. 2, April 1999.
Real-Time Simulation of Dust BehaviorFig. 1. 85The applications of particle systems.Structured particle systems are often used to model trees, water drops,leaves, grass, rainbow, and clouds. Stochastic particle systems are oftenused to model fireworks, explosions, snow, and so forth. Oriented particlesystems are often used to model deformable and rigid bodies such as cloth,lava flow, etc.Figure 2 is an example of using particle systems to model grass from theCenter for Computational Statistics, George Mason University.Dust behavior is fuzzy and unstructured and, thus, belongs to thecategory of stochastic particle systems as indicated in Figure 1. In thefollowing section, we construct a stochastic particle system to simulate dustbehavior caused by a fast-traveling vehicle.3. DUST BEHAVIORSIn this section, we introduce the methods for the generation of dustparticles, discuss the dynamics of air flow around the vehicle that affect thedust behavior, and analyze the forces acting on a dust particle in order toestablish corresponding physically based empirical models.3.1 Dust GenerationThere are many factors affecting the generation of dust particles (numberof particles, initial locations, and velocities) as well as affecting the behavior of the particles after their generation. Here we develop correlationsamong the factors and the dust generation process, so that when we have adifferent vehicle or the same vehicle in a different environment, we canchange the simulations just by adjusting the corresponding parameters.As a vehicle wheel passes over an unpaved surface with velocity Vcar!vertical pressure, due to the weight of the vehicle WTcar! , will produceground surface vibration and deformation, crushing large particles intosmaller ones, and splashing particles into the air (Figure 3(a), left). WT carACM Transactions on Modeling and Computer Simulation, Vol. 9, No. 2, April 1999.
86 J. X. Chen et al.Fig. 2.Grass.and Vcar both affect the number of particles generated. Vcar also affects theinitial velocity of the particle generated at the bottom of the tire.Horizontal stress and friction are largely due to the driving power thatsustains the velocity Vcar! and acceleration of the vehicle and will furtherpulverize the particles and carry them on the surface of the tire. Theslippage between the tire and the ground surface will lift particles ofdifferent sizes due to the adhesive and shear forces, and eject them atdifferent places on the tire surface due to the centrifugal force and air dragforce (Figure 3(a), right). Vcar decides the initial speed of the particle on thetire surface. Angular velocity of the tire v tire! , radius of the tire Rtire ! ,particle volume or size qp! , particle density rp! , and particle stickiness sp! affect the location of the particle leaving the tire surface.The pressure gradient formed by the moving air underneath and behindthe vehicle will also lift fine particles (Figure 3(b)). The initial velocity ofthe particle depends on the energy associated with the pressure gradient inthe vertical direction. The length Lcar! , height Hcar! , and width Wcar! ofthe vehicle are used together to calculate the pressures and velocities in thefluid region (introduced in the next section).There are many other important factors associated with the conditions ofthe environment that affect dust generation: the environmental wind Vwd!,the density of dust on the ground r gd ! , the average volume or size ofparticles on the ground qgd! , and the adhesion and wetness of the ground sgd! . If the ground is wet and the average size of the particles is large,there will be fewer particles. If the dust density of the ground surface ishigh, there will be more particles.ACM Transactions on Modeling and Computer Simulation, Vol. 9, No. 2, April 1999.
Real-Time Simulation of Dust BehaviorTire 87Tire adhesive and shear forcescounter the centrifugal,gravity, and air drag ss & frictionVertical Pressurea) Particles generated around the tireBody of the vehiclePhhWheelDrivingDirectionPob) Particles generated underneath and behind the vehicleFig. 3.Forces between the wheels and the ground.In summary, most dust particles are generated right behind and alongside of the wheels. Some fine dust particles are lifted from the groundsurface because of the pressure gradient underneath and behind thevehicle. Each particle is generated with its own initial density, size,position, and velocity. In addition to the parameters that are used to affectthe dust generation as listed in Table 1, a particle’s shape affects thedirection of its velocity [Wejchert and Haumann 1991], which is included asrandom perturbation of the velocity. Particle mass and size are randomlygenerated with a Poisson distribution.After a particle is generated with its mass mp 5 rpqp! , stickiness, initialposition and velocity, it is then entrained in the turbulent air behind thevehicle. It will eventually return to the ground depending on its propertiesand environment conditions. Bigger particles will fall back to the groundsurface more rapidly while the fine ones will remain suspended in the air.Small stones and blocks of muds may fall back to the ground immediatelyafter ejection from the tires. The turbulent air around and behind thevehicle is calculated according to fluid dynamics as described in the nextsection.ACM Transactions on Modeling and Computer Simulation, Vol. 9, No. 2, April 1999.
88J. X. Chen et al. Table I.ItemsVehicleParameters affecting the dust generationParametersDescriptionVelocity - VcarDecide the velocity of thefluid and boundary of the carwhere the turbulence isgenerated; determine thepressures and velocities of thefluid field;Decide where a dust particleleaves the tireWeight - WT carTire radius - R tireHeight - H carWidth - W carLength - L carDust ParticleVolume (size) - q pDensity - r pStickiness - s pShape (included as randomperturbation of velocities)Decide where a dust particleleaves the tire and ground;affect the number of particlesgeneratedEnvironmentGround density - r gdAverage Volume (size) - q gdGround wetness - s gdEnvironmental Wind - VwdInfluence the number ofparticles generated and theirinitial velocities.3.2 Fluid Dynamics Affecting Dust BehaviorsThe physical problem of the air flow around a moving vehicle belongs to thewell-studied subject of aeromechanics which is called separated flow [Burggraf 1966; Prandtl and Tietjens 1934]. Burggraf [1966] summarized thevarious regimes of flow experienced by circular cylinder in an incompressible fluid (Figure 4). The appearance of a wake first occurs at ReynoldsNumber 1 Re! of the order of 1, and the flow separates from the rear of thecylinder, forming a recirculating eddy for Reynolds number greater than 5;the steady recirculating eddy persists at Reynolds number about the orderof 100; the steady flow then breaks down into the Karman vortex street;finally the stream develops into a completely turbulent wake for Reynoldsnumber greater than the order of 10 5 .Considering a vehicle at speed Vcar 5 60 km/hour, vehicle size of theorder of L 5 1 meter, and an ordinary air kinematics viscosity of 10 25 ,the Reynolds numbers Re! at the boundary and right behind the vehicleare then in the order of 10 6 . Therefore, turbulence is the main feature ofthe flow at the boundary of the vehicle movement in our model. The studyof turbulence is still not satisfactory [Burggraf 1966]. The situation isworse in simulation with numerical models: even the most advancedmodels of today, such as direct numerical simulation (DNS), large eddysimulation (LES), or Reynolds average equations (RANS), are deficient forlarge Reynolds numbers [Gatski 1996]. As for graphics simulation applicaRe 5 VL / v where L and V are characteristic length and velocity respectively, and y is thekinematic viscosity. For example, given fluid flow inside a pipe, then L can be the diameter ofthe pipe and V the velocity of the fluid flow.1ACM Transactions on Modeling and Computer Simulation, Vol. 9, No. 2, April 1999.
Real-Time Simulation of Dust BehaviorReynoldsnumber rangeFlow configuration89RemarksRe 1Steady flow, no wakeRe 1Steady flow, non-circulation wakeRe 102Relatively steady, recirculating flowRe 103Stable vortex streetRe 105Turbulent wakeFig. 4. Flow regimes for circular cylinder (from Burggraf [1966]).tions, it is not necessary to simulate turbulence in such detail, which isalmost random motion in higher order scales. Therefore, in this paper, weassume that the flow around a moving vehicle can be approximated as asteady, low Reynolds external laminar flow modified by a random velocityin the high Reynolds turbulent region. Here we choose the separationbetween the turbulent flow and the laminar flow by the tangential
ior and achieving simulation of the behavior in real time. Categories and Subject Descriptors: I.6.5 [Simulation and Modeling]: Model Development— Modeling methodologies General Terms: Simulation, Graphics, Dust Additional Key Words and Phrases: Physically-based Modeling, Real-time Simulation, Vehicle, Particle Systems, Computational Fluid .
month and year of start and end for each occupation. We also asked about occupational exposures during the past 12 months, i.e., gas, smoke or dust; smell of frying; car exhaust fumes/engine fumes; mineral dust; organic dust (flour dust, wood dust, paper dust or textile dust); inorganic dust (grinding, milling, turning, mineral wool,
DUST COLLECTOR The Model CC Cyclone separator, dust collector blows the dust laden air into the long tapered cyclone by means of a powerful radial blade blower. The high velocity cyclone action separates the dust particles into a dust storage cabinet or a 55- gallon drum. Dust free air is then returned to the work place. For very ne dust an after
Dust emissions can be prevented or reduced in four basic ways: 1. Limit the creation or presence of dust-sized particles. 2. Reduce wind speed at ground level. 3. Bind dust particles together. 4. Capture and remove dust from its sources. Some dust control techniques work with many sources of dust:
visible dust emissions, and in the case of dust emission from inactive disturbed areas during high winds additional watering or dust suppressants shall be applied to reduce the visible dust emissions. Bulk material storage piles shall be covered or stored in areas with wind barriers and water/dust suppressants applied to reduce dust emissions.
Transients (EMT), power systems modeling & simulation, and . First, real-time simulation is defined. An overview of the evolution of real-time simulators is then presented. Two other essential questions are then answered. Why is real-time simulation needed? Where does real-time simulation fit best? Finally, this paper concludes
Super Dust Deputy / Super Dust Deputy XL Super Dust Deputy Deluxe 4” Owner’s Manual AXD002030A AXD002040A AXD002030AXL Made in the USA ZBM000036SDDXLA / 1-2017 1.800.732.4065 www.oneida-air.com.5
ism [15] and a visual Agent simulation based on a 3D real-time visualization simulation in real-time. The visual agent simulation allows us to visualize the land combat simula-tion scenarios in a 3D scene. We also propose to reduce the programming time consumed to develop or modify the sce-nario tactics for combat in real-time visual simulation by us-ing Cell-DEVS, which allows us to use .
Diversity of Algae, Lichens, & Bryophytes 50 Paper III Diversity of Pteridophytes & Gymnosperms 50 Practical Practical Syllabus based on theory papers 50 B.Sc. II Year Papers Title of Paper Max. Marks Paper I Diversity of Angiosperms: Systematics, Development & Reproduction 50 Paper II Cytology, Genetics, Evolution & Ecology 50 Paper III Plant Physiology and Biochemistry 50 Practical Practical .
